<!--[if gte mso 9]> Normal 0 false false false MicrosoftInternetExplorer4 <!-- /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:""; margin:0in; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:"Times New Roman"; mso-fareast-font-family:"Times New Roman";} @page Section1 {size:8.5in 11.0in; margin:1.0in 1.25in 1.0in 1.25in; mso-header-margin:.5in; mso-footer-margin:.5in; mso-paper-source:0;} div.Section1 {page:Section1;} --> <!--[if gte mso 10]> I am planning to install a hydronic radiant system under two bedrooms in our 1953 brick ranch in Atlanta with an unheated crawlspace. 

 

I would go with 8” spacing and use some lightweight aluminum plates, but the joists are on 12” centers.  I could go with 2 runs in every bay, but that seems like overkill.  Will I be able to bend the pex to such a tight curve?  How much difference will the lightweight plates make compared to no plates?  I really can’t justify the extra cost of the heavy plates.  We are talking Atlanta weather, which is usually pretty mild (This past week is an exception).  I think that someone has stated that a radiant barrier is unnecessary under the system when plates are used.  Is this true?  I will use r-19, but do I need foil batts?  If so, does anyone know any manufacturer that makes a 24” foil faced back?  (With 12” joist spacing I have cut down insulation to fit in the past)  A 12” foil batt would be even better.

 

I think my heat load is about 10000-12000 BTU/hr total for the two rooms.  One room is unrenovated (original windows with storm windows with no wall insulation beside brick) and the other has low-e double pane windows and r-13 in walls.  Each room is about 170 square ft.

 

I will heat the water with a gas hot water heater.  Yes, I know there are mixed opinions about this, but once again I am talking Atlanta.

 

All advice and opinions are welcome.

atlanta or not, if those heat loads are correct, be careful. you're at 30 BTUs/sq ft there which is close to the max for any radiant floor system under wood.

I imagine the vast majority of the load is in the unrenovoated room and that it won't keep up even with heavy plates.

a water heater is an excellent choice for the load, but if you need 180 degree water here it won't do what you need it to do.

strongly suggest renovating that other room first. otherwise you need to decide what your supplemental heat is going to be.

you may not need lightweight plates at all in the other room.

use two runs per joist bay.

dont' need foil batts. anything opaque (kraft paper) should do the trick.

Radiant barrier would have extremely marginal value for a plated system, where most of the heat transfer from the tubing to the floor is conducted via the aluminum plate. The emissivity of aluminum is already very low, and since it's running at low temp even if it weren't the downward radiated heat transfer would still be pretty low. The plates being aluminum it goes from "pretty low" to "extremely low".

On suspended-tube (unplated), there's some rationale for radiant barrier, but simply going to high density "cathedral ceiling" batts (R25-ish instead of R19-ish). Alternatively, using 1/2" OSB or plywood screwed in from below (taped seams, and foam sealed at all ends or penetrations to prevent convection) and blowing 5-6" of cellulose into the cavities (might be difficult to measure) puts a foil vapor retarder on the high-humidity side and you could avoid the misery of cutting down fiberglass batts. You might use housewrap stapled below the tubing to keep the blown insulation from covering the tubing and define the depth if it's suspended-tube. Housewrap is highly water-vapor permeable, and would allow the fiber to dry toward the air-conditioned interior. OSB is semi-permeable- but an order of magnitude less permeable than housewrap. (I think this is how I'd go even for a plated portion rather than cutting down batts.)

Over a crawlspace in a cooling dominated high-humidity environment like Atlanta, if the crawlspace is freely vented to the outside, the least-permeable face needs to be on the crawlspace side of the insulation, putting the fiber in same hygric space as the conditioned-space side. Otherwise warm humid air from the crawlspace can rise through the fiber to condense in the fiber or on the joists. But this can be difficult to figure at a distance. A better solution overall would be to seal & insulate the crawlspace walls, put a vapor retarder on the crawlspace floor, use unfaced insulation under the radiant, and have a small amount of summertime AC flowing through the crawlspace (sometimes just a dehumidifier will do). This is a whole separate project which, could be difficult to implement, depending on the particulars, but may be the right thing to do from both a heating and AC point of view. But were you to use foil or kraft facers they would have to be on the conditioned space for the radiant barrier effect or kraft facers for the desired effect in the heating season, which would be the WRONG side for the most of the year, and mold issues would be all but certain. (Mind you, in heating dominated climates you would never want the facers to be on the crawlspace side for similar mold-hazard reasons.)

So the stackup you probably want is:

[conditioned space]

Blowing cellulose isn't rocket science, and you don't need to dense-pack it, any box-store rental unit and a 2-hole method would be plenty for this application. Seal the blowing holes with aluminum FSK tape or spray foam when you're done.

Floor
subfloor
tubing
housewrap (highly permeable)
cellulose/fiberglass (R15 min, R20 preferable)
OSB/plywood

[crawlspace]

Thanks for the responses. Rob, I agree that renovating the one room is a priority. What did you use for your temperatures (desired ambient and outdoor) to achieve your 180 degree water temp? I am really looking for a way to keep the floors warm (~75 F) until I can get around to gutting my son's room. I was pretty conservative with the heat load. We really only have 5-7 days a winter below 20 F on average.   What should be my heat load goal in btu/sq ft to achieve a 140 F water temperature?

Dana, wow! Thanks for that thorough insulation explanation. Crawlspace encapsulation is definitely on my short list, but I have section of the crawlspace with a skimcoat concrete floor workshop/storage that will have to be torn out to vapor seal and address and intermittent rainwater leak. If it wasn't for that, the encapsulation would be done by now. Atlanta is tricky with the mostly heating or cooling. Most people consider it a mostly heating environment and keep the vapor barrier on the inside of the walls.

I forgot to mention that the subfloor is 1x6" lumber with 1/16"-1/8" between each board topped with felt paper and 3/4" oak floors. Does this help or hurt me with the underfloor installation?

outdoor is not relevant when you give me a 30 BTU per square foot load (that's taken into account already if a heat load has been done). light plates can only put out that much heat with about 180 degree water... IF it could it then. if you drop your room temp from 70 you can drop your water temp the same number of degrees. 140 with light plates with regular subfloor and 1 wood floor layer on top won't break 20 BTUs/sq ft. In your case, slightly less with the thicker floor.

Atlanta is definitely a mixed climate, and the heating degree-days DO outnumber the cooling degree days, but the crawlspace/floor is still a different set of environmental conditions than an exterior wall.  While unconditioned crawlspace humidity will rise & fall with outdoor enthalpy, it's will be reliably HIGHER than outdoor air.  If at the same time crawlspace temperature is also significantly higher than the conditioned space above, air infiltration from the crawlspace must be reliably blocked, and vapor retardency must be higher on the warm-wet side. Latent loads are quite high in GA- this wouldn't be an issue at all in similarly hot places with drier climates. (Like I said, it's difficult to analyze the particulars on this one at a distance.)

But if the long term plan is to seal & insulate the crawlspace, making it part of the conditioned space and you don't need to sweat the vapor retardency issues of the floor stackup, and you can back off to R10-R13 (3-4" of cellulose) on the under-radiant insulation without experiencing a big bump in the heating bill.  Short term, painting the non-structural slab with a radon-sealer or vapor retardent paint can cut the ground vapor intrusion by a large fraction, and running a dehumidifier to keep it at or under 60% RH will ensure you don't run into mold problems.

I think I am going to go with some thinfin-c plates. Hopefully that will bring down my water temps. I am also thinking of putting in 2 runs and every other bay and 1 in the others. That will put my spacing at 7",8.5",7",8.5" ... Will this cause any noticeable variation in floor temps? How will this affect my performance or water temps?

Is there formula or calculation method for determining my delta-T. I am familiar with this formula: f = Q/(500 x delta-T). If my design day output is 18 btu/sq ft, that would put a 150 ft loop at 1700 Btu/hr. Since all 1700 Btu/hr will not go into the room. How do design for a delta-T of less that 15-20 F with a flow rate of 0.9 gpm. Basically, how much btu/sq ft should I use in the formula to get my desired output with my staple situation and method, or is there just a better way to go about this using the loop length and flow rate?

do not skip runs with the plates. you want 2 runs per bay or the install is a bear and if your load is as high as you say it was (30 BTUs/sqft.. but now you say 18?) you should not be cutting corners if you want to get your water temps down.

I don't know what you are trying to do with flow. You use load and delta-T to determine your flow rate requirements. I recommend between 10 and 20 degree dt, with 10 for maximum supply temp limitation and 20 for flow economy.. there is a 5 degree difference in Supply water requirement between those two figures.

Rob, I must admit my post is confusing. First, I have decided to go ahead with the room gutting/renovation this summer. I think I can get the heat load down to 3000-3300 btu/hr with the renovation. That brings me down to about 18-19 btu/sq ft for each room. Forget about the 0.9 flow rate. So if I have two loops for each room, each loop should provide ~1600 BTU/hr. Therefore f=1600/500/15=0.21 gpm. Using "ft of head=0.0374*1*190 ft*(0.21)^1.75", if I have a 190 ft run that means I need a pump with 0.5 ft of head. If I use one run of 380 ft (f =3200/500/15=.42 gpm) that would be a pump with about 2 ft of head. What is the minimum flow rate for 1/2" pex anyway? Why is it that everyone states ~300 ft max runs for 1/2" pex. It seems like I could run the 380 ft loops with no problem. I could run 12-13 380 ft loops on a Taco 005. Rob, please correct with appropriate math if I am wrong.

No back to the original question. According to the graph at http://www.radiantengineering.com/ThermoFin/ThermofinOutput.png. Thermofin (not the thinfin c) will put out 20 btu/sq ft with an R=1 on 8" OC. The thinfin will obviously put out less than this. I need to see if RE has that data. Rob, maybe you already have it. Anyway, why is it that going with an average center of 7.75" will not work for my design. If you think that the install difficulties is the problem, I already have a joist block in the middle of each bay. How does going through the other end of the joist complicate things? I am thinking of using thermapex which is pex-a and more flexible and may be better suited to making the tight bends in the 12" cavity. Will this help or am I just wasting money with a higher quality pex?

Thanks so much for you input.

two runs for bays allows you to not have to thread every foot of pipe through every hole, you can instead use the "pull method" which is much, much, much easier. in that case I would at least do dual runs of pipe in every joist bay even if they aren't both plated, at an absolute bare minimum. the "extra" run could even just hang there for all I would care, as long as it's there. the cost of the pipe would be more than offset in labor reduction.

if you have 12" o.c. joists, you'd basically be 6", 9", 9", 6". I think you would have noticeable striping in that configuration. less with the thicker thermofin, but still, I think it would be noticeable where the 6"s are vs the 9". Dunno where you got your spacing but it has to equal 24" for 3 runs if you're going to do a regular spacing.

Thanks Rob,
I got my spacing by pushing the 3.5" thinfin up against the joist. I assume that there is some reason I can't do that. I agree, the pull method is simplest. I suppose since each bay has a center block (from subfloor to bottom of joist), I can pull from both ends of each bay and loop back at the block.

I came up with the reduce spacing to reduce the cost of plates not tubing. I will just run 2 runs per bay, and hang the second run

Now about my flow rate question. Can I run 1/2" pex as low as 0.21 or 0.42 gpm? If not, how low of a flow rate can I go? How long can I go on a run of 1/2" pex? With a flow rate of 0.42 it seems that I can go well over 350.

if you push a 3.5" thinfin against a 2x joist, you are 8.5", 7", 8.5"... sorry, I misinterpreted what you posted somehow, you're right there. That should be ok, I guess. I would call it close enough to 8" o.c.

we run 0.2 GPM all the time, but it depends on load and temperature drop what you can run. at 18 BTUs/sq ft and a 20 degree Dt that's something like 110 sq ft or a loop length of something like 180 in your case.